The Linux kernel’s ext4 filesystem contains a retry loop in ext4_xattr_block_set() that can be triggered when a cache entry’s reference count is already saturated. The routine retries the lookup without marking that entry unusable, allowing the same saturated entry to be retried indefinitely. A task that spins in this loop can hold the parent directory’s write lock, causing concurrent rmdir callers to block and effectively hanging filesystem activity. This flaw is triggered when on‑disk metadata violates the invariant that a reusable entry has a count below EXT4_XATTR_REFCOUNT_MAX, for example through corruption or an overflow condition.

All Linux kernel installations that have not incorporated the patch to the ext4_xattr code path are affected. The vendor and product are the Linux kernel. No specific distribution or kernel version is listed in the CVE entry, so virtually any Linux kernel variant before the patched commit is vulnerable.

The issue lacks a CVSS or EPSS score and is not listed in the CISA KEV catalog, indicating limited publicly documented exploitation. The exploit requires local or privileged access to corrupt ext4 metadata or induce an overflow condition. Once triggered, a hung task blocks directory removal operations and can stall system activity, but it does not provide remote code execution or data exfiltration. The risk is considered moderate to high for local attackers with the necessary filesystem access, while the external threat remains limited under current data.